{"id":"W2907670343","doi":"10.48550/arxiv.1903.03088","title":"Self-Tuning Networks: Bilevel Optimization of Hyperparameters using Structured Best-Response Functions","year":2019,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":29,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Hyperparameter; Hyperparameter optimization; Computer science; Artificial intelligence; Artificial neural network; Machine learning; Mathematical optimization; Mathematics; Support vector machine","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001150966,0.0001463984,0.0001694261,0.0001828076,0.0001574588,0.00003035623,0.0006103356,0.00007959965,0.00001811323],"category_scores_gemma":[0.0000198819,0.0001690043,0.0000791163,0.00146403,0.00005247058,0.0007043511,0.0002166473,0.0001334582,0.00001816076],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001159251,"about_ca_system_score_gemma":0.00006667426,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000008538767,"about_ca_topic_score_gemma":0.000002109961,"domain_scores_codex":[0.9989205,0.0001153718,0.0001656594,0.0004873046,0.00006733538,0.0002438552],"domain_scores_gemma":[0.9986079,0.0002487266,0.000200058,0.0007243759,0.0001278119,0.00009117343],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005135289,0.00003255208,0.001511382,0.000004290165,0.00002098222,0.000004141697,0.00004443607,0.9899935,0.001062485,0.006861475,0.000006442408,0.0004069499],"study_design_scores_gemma":[0.0003632904,0.0000672414,0.0002727562,0.00001827906,0.00002862482,0.000006929103,0.00005264928,0.9980452,0.0001633295,0.00067695,0.0001317524,0.0001729967],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3402303,0.0000119668,0.6591909,0.00001465257,0.0001247285,0.0002516731,0.000002235645,0.0001105799,0.00006295089],"genre_scores_gemma":[0.8517873,0.00001563607,0.1479435,0.00003775126,0.00001725836,8.231941e-7,0.000003142271,0.00001221186,0.0001824402],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.511557,"threshold_uncertainty_score":0.6891794,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04801974787008525,"score_gpt":0.1928887671788964,"score_spread":0.1448690193088112,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}